Transcriptional landscape estimation from tiling array data using a model of signal shift and drift

Staphylococcus aureus Transcriptome Architecture: From Laboratory to Infection-Mimicking Conditions

Staphylococcus aureus is a major pathogen that colonizes about 20% of the human population. Intriguingly, this Gram-positive bacterium can survive and thrive under a wide range of different conditions, both inside and outside the human body. Here, we investigated the transcriptional adaptation of S. aureus HG001, a derivative of strain NCTC 8325, across experimental conditions ranging from optimal growth in vitro to intracellular growth in host cells. These data establish an extensive repertoire of transcription units and non-coding RNAs, a classification of 1412 promoters according to their dependence on the RNA polymerase sigma factors SigA or SigB, and allow identification of new potential targets for several known transcription factors. In particular, this study revealed a relatively low abundance of antisense RNAs in S. aureus, where they overlap only 6% of the coding genes, and only 19 antisense RNAs not co-transcribed with other genes were found. Promoter analysis and comparison with Bacillus subtilis links the small number of antisense RNAs to a less profound impact of alternative sigma factors in S. aureus. Furthermore, we revealed that Rho-dependent transcription termination suppresses pervasive antisense transcription, presumably originating from abundant spurious transcription initiation in this A+T-rich genome, which would otherwise affect expression of the overlapped genes. In summary, our study provides genome-wide information on transcriptional regulation and non-coding RNAs in S. aureus as well as new insights into the biological function of Rho and the implications of spurious transcription in bacteria.

The major human pathogen Staphylococcus aureus can survive under a wide range of conditions, both inside and outside the human body. The goal of this study was to determine how S. aureus adapts to such different conditions and, additionally, we wanted to identify general factors governing the staphylococcal transcriptome architecture. Therefore, we performed a precise analysis of all RNA transcripts of S. aureus across experimental conditions ranging from in vitro growth in different media to internalization by eukaryotic host cells. We systematically mapped all transcription units, annotated non-coding RNAs, and assigned promoters controlled by particular RNA polymerase sigma factors and transcription factors. By a comparison with data available for the related Gram-positive bacterium Bacillus subtilis, we made key observations concerning the abundance and origin of antisense RNAs. Intriguingly, these findings support the view that many antisense RNAs in a bacterium like B. subtilis could be byproducts of spurious promoter recognition by condition-specific alternative sigma factors. We also report that the transcription termination factor Rho prevents widespread antisense transcription, presumably caused by pervasive transcription initiation in the A+T-rich genome of S. aureus. Altogether our study presents new perspectives on the biological significance of antisense and pervasive transcription in bacteria.

Transcript mapping based on dRNA-seq data

Background

RNA-seq and its variant differential RNA-seq (dRNA-seq) are today routine methods for transcriptome analysis in bacteria. While expression profiling and transcriptional start site prediction are standard tasks today, the problem of identifying transcriptional units in a genome-wide fashion is still not solved for prokaryotic systems.

Results

We present RNAseg, an algorithm for the prediction of transcriptional units based on dRNA-seq data. A key feature of the algorithm is that, based on the data, it distinguishes between transcribed and un-transcribed genomic segments. Furthermore, the program provides many different predictions in a single run, which can be used to infer the significance of transcriptional units in a consensus procedure. We show the performance of our method based on a well-studied dRNA-seq data set for Helicobacter pylori.

Conclusions

With our algorithm it is possible to identify operons and 5’- and 3’-UTRs in an automated fashion. This alleviates the need for labour intensive manual inspection and enables large-scale studies in the area of comparative transcriptomics.

Parseq: reconstruction of microbial transcription landscape from RNA-Seq read counts using state-space models

MOTIVATION

The most common RNA-Seq strategy consists of random shearing, amplification and high-throughput sequencing of the RNA fraction. Methods to analyze transcription level variations along the genome from the read count profiles generated by the RNA-Seq protocol are needed.

RESULTS

We developed a statistical approach to estimate the local transcription levels and to identify transcript borders. This transcriptional landscape reconstruction relies on a state-space model to describe transcription level variations in terms of abrupt shifts and more progressive drifts. A new emission model is introduced to capture not only the read count variance inside a transcript but also its short-range autocorrelation and the fraction of positions with zero counts. The estimation relies on a particle Gibbs algorithm whose running time makes it more suited to microbial genomes. The approach outperformed read-overlapping strategies on synthetic and real microbial datasets.

AVAILABILITY

A program named Parseq is available at: http://www.lgm.upmc.fr/parseq/.

CONTACT

bodgan.mirauta@upmc.fr

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Detecting splicing variants in idiopathic pulmonary fibrosis from non-differentially expressed genes

Idiopathic pulmonary fibrosis (IPF) is an interstitial lung disease of unknown cause that lacks a proven therapy for altering its high mortality rate. Microarrays have been employed to investigate the pathogenesis of IPF, but are presented mostly at the gene-expression level due to technologic limitations. In as much as, alternative RNA splicing isoforms are increasingly identified as potential regulators of human diseases, including IPF, we propose a new approach with the capacity to detect splicing variants using RNA-seq data. We conducted a joint analysis of differential expression and differential splicing on annotated human genes and isoforms, and identified 122 non-differentially expressed genes with a high degree of "switch" between major and minor isoforms. Three cases with variant mechanisms for alternative splicing were validated using qRT-PCR, among the group of genes in which expression was not significantly changed at the gene level. We also identified 35 novel transcripts that were unique to the fibrotic lungs using exon-exon junction evidence, and selected a representative for qRT-PCR validation. The results of our study are likely to provide new insight into the pathogenesis of pulmonary fibrosis and may eventuate in new treatment targets.

Bacillus subtilis RNase Y activity in vivo analysed by tiling microarrays

RNase Y is a key endoribonuclease affecting global mRNA stability in Bacillus subtilis. Its characterization provided the first evidence that endonucleolytic cleavage plays a major role in the mRNA metabolism of this organism. RNase Y shares important functional features with the RNA decay initiating RNase E from Escherichia coli, notably a similar cleavage specificity and a preference for 5′ monophosphorylated substrates. We used high-resolution tiling arrays to analyze the effect of RNase Y depletion on RNA abundance covering the entire genome. The data confirm that this endoribonuclease plays a key role in initiating the decay of a large number of mRNAs as well as non coding RNAs. The downstream cleavage products are likely to be degraded by the 5′ exonucleolytic activity of RNases J1/J2 as we show for a specific case. Comparison of the data with that of two other recent studies revealed very significant differences. About two thirds of the mRNAs upregulated following RNase Y depletion were different when compared to either one of these studies and only about 10% were in common in all three studies. This highlights that experimental conditions and data analysis play an important role in identifying RNase Y substrates by global transcriptional profiling. Our data confirmed already known RNase Y substrates and due to the precision and reproducibility of the profiles allow an exceptionally detailed view of the turnover of hundreds of new RNA substrates.

Analysis of tiling array expression studies with flexible designs in Bioconductor (waveTiling)

BACKGROUND

Existing statistical methods for tiling array transcriptome data either focus on transcript discovery in one biological or experimental condition or on the detection of differential expression between two conditions. Increasingly often, however, biologists are interested in time-course studies, studies with more than two conditions or even multiple-factor studies. As these studies are currently analyzed with the traditional microarray analysis techniques, they do not exploit the genome-wide nature of tiling array data to its full potential.

RESULTS

We present an R Bioconductor package, waveTiling, which implements a wavelet-based model for analyzing transcriptome data and extends it towards more complex experimental designs. With waveTiling the user is able to discover (1) group-wise expressed regions, (2) differentially expressed regions between any two groups in single-factor studies and in (3) multifactorial designs. Moreover, for time-course experiments it is also possible to detect (4) linear time effects and (5) a circadian rhythm of transcripts. By considering the expression values of the individual tiling probes as a function of genomic position, effect regions can be detected regardless of existing annotation. Three case studies with different experimental set-ups illustrate the use and the flexibility of the model-based transcriptome analysis.

CONCLUSIONS

The waveTiling package provides the user with a convenient tool for the analysis of tiling array trancriptome data for a multitude of experimental set-ups. Regardless of the study design, the probe-wise analysis allows for the detection of transcriptional effects in both exonic, intronic and intergenic regions, without prior consultation of existing annotation.

Wavelet-based detection of transcriptional activity on a novel Staphylococcus aureus tiling microarray

BACKGROUND

High-density oligonucleotide microarray is an appropriate technology for genomic analysis, and is particulary useful in the generation of transcriptional maps, ChIP-on-chip studies and re-sequencing of the genome.Transcriptome analysis of tiling microarray data facilitates the discovery of novel transcripts and the assessment of differential expression in diverse experimental conditions. Although new technologies such as next-generation sequencing have appeared, microarrays might still be useful for the study of small genomes or for the analysis of genomic regions with custom microarrays due to their lower price and good accuracy in expression quantification.

RESULTS

Here, we propose a novel wavelet-based method, named ZCL (zero-crossing lines), for the combined denoising and segmentation of tiling signals. The denoising is performed with the classical SUREshrink method and the detection of transcriptionally active regions is based on the computation of the Continuous Wavelet Transform (CWT). In particular, the detection of the transitions is implemented as the thresholding of the zero-crossing lines. The algorithm described has been applied to the public Saccharomyces cerevisiae dataset and it has been compared with two well-known algorithms: pseudo-median sliding window (PMSW) and the structural change model (SCM). As a proof-of-principle, we applied the ZCL algorithm to the analysis of the custom tiling microarray hybridization results of a S. aureus mutant deficient in the sigma B transcription factor. The challenge was to identify those transcripts whose expression decreases in the absence of sigma B.

CONCLUSIONS

The proposed method archives the best performance in terms of positive predictive value (PPV) while its sensitivity is similar to the other algorithms used for the comparison. The computation time needed to process the transcriptional signals is low as compared with model-based methods and in the same range to those based on the use of filters. Automatic parameter selection has been incorporated and moreover, it can be easily adapted to a parallel implementation. We can conclude that the proposed method is well suited for the analysis of tiling signals, in which transcriptional activity is often hidden in the noise. Finally, the quantification and differential expression analysis of S. aureus dataset have demonstrated the valuable utility of this novel device to the biological analysis of the S. aureus transcriptome.

Genome-wide identification of genes directly regulated by the pleiotropic transcription factor Spx in Bacillus subtilis

The transcriptional regulator Spx plays a key role in maintaining the redox homeostasis of Bacillus subtilis cells exposed to disulfide stress. Defects in Spx were previously shown to lead to differential expression of numerous genes but direct and indirect regulatory effects could not be distinguished. Here we identified 283 discrete chromosomal sites potentially bound by the Spx–RNA polymerase (Spx–RNAP) complex using chromatin immunoprecipitation of Spx. Three quarters of these sites were located near Sigma(A)-dependent promoters, and upon diamide treatment, the fraction of the Spx–RNAP complex increased in parallel with the number and occupancy of DNA sites. Correlation of Spx–RNAP-binding sites with gene differential expression in wild-type and Δspx strains exposed or not to diamide revealed that 144 transcription units comprising 275 genes were potentially under direct Spx regulation. Spx-controlled promoters exhibited an extended −35 box in which nucleotide composition at the −43/−44 positions strongly correlated with observed activation. In vitro transcription confirmed activation by oxidized Spx of seven newly identified promoters, of which one was also activated by reduced Spx. Our study globally characterized the Spx regulatory network, revealing its role in the basal expression of some genes and its complex interplay with other stress responses.

13C-flux analysis reveals NADPH-balancing transhydrogenation cycles in stationary phase of nitrogen-starving Bacillus subtilis

In their natural habitat, microorganisms are typically confronted with nutritional limitations that restrict growth and force them to persevere in a stationary phase. Despite the importance of this phase, little is known about the metabolic state(s) that sustains it. Here, we investigate metabolically active but non-growing Bacillus subtilis during nitrogen starvation. In the absence of biomass formation as the major NADPH sink, the intracellular flux distribution in these resting B. subtilis reveals a large apparent catabolic NADPH overproduction of 5.0 ± 0.6 mmol g(-1)h(-1) that was partly caused by high pentose phosphate pathway fluxes. Combining transcriptome analysis, stationary (13)C-flux analysis in metabolic deletion mutants, (2)H-labeling experiments, and kinetic flux profiling, we demonstrate that about half of the catabolic excess NADPH is oxidized by two transhydrogenation cycles, i.e. isoenzyme pairs of dehydrogenases with different cofactor specificities that operate in reverse directions. These transhydrogenation cycles were constituted by the combined activities of the glyceraldehyde 3-phosphate dehydrogenases GapA/GapB and the malic enzymes MalS/YtsJ. At least an additional 6% of the overproduced NADPH is reoxidized by continuous cycling between ana- and catabolism of glutamate. Furthermore, in vitro enzyme data show that a not yet identified transhydrogenase could potentially reoxidize ∼20% of the overproduced NADPH. Overall, we demonstrate the interplay between several metabolic mechanisms that concertedly enable network-wide NADPH homeostasis under conditions of high catabolic NADPH production in the absence of cell growth in B. subtilis.

Condition-dependent transcriptome reveals high-level regulatory architecture in Bacillus subtilis

Bacteria adapt to environmental stimuli by adjusting their transcriptomes in a complex manner, the full potential of which has yet to be established for any individual bacterial species. Here, we report the transcriptomes of Bacillus subtilis exposed to a wide range of environmental and nutritional conditions that the organism might encounter in nature. We comprehensively mapped transcription units (TUs) and grouped 2935 promoters into regulons controlled by various RNA polymerase sigma factors, accounting for ~66% of the observed variance in transcriptional activity. This global classification of promoters and detailed description of TUs revealed that a large proportion of the detected antisense RNAs arose from potentially spurious transcription initiation by alternative sigma factors and from imperfect control of transcription termination.

Three essential ribonucleases-RNase Y, J1, and III-control the abundance of a majority of Bacillus subtilis mRNAs

Bacillus subtilis possesses three essential enzymes thought to be involved in mRNA decay to varying degrees, namely RNase Y, RNase J1, and RNase III. Using recently developed high-resolution tiling arrays, we examined the effect of depletion of each of these enzymes on RNA abundance over the whole genome. The data are consistent with a model in which the degradation of a significant number of transcripts is dependent on endonucleolytic cleavage by RNase Y, followed by degradation of the downstream fragment by the 5′–3′ exoribonuclease RNase J1. However, many full-size transcripts also accumulate under conditions of RNase J1 insufficiency, compatible with a model whereby RNase J1 degrades transcripts either directly from the 5′ end or very close to it. Although the abundance of a large number of transcripts was altered by depletion of RNase III, this appears to result primarily from indirect transcriptional effects. Lastly, RNase depletion led to the stabilization of many low-abundance potential regulatory RNAs, both in intergenic regions and in the antisense orientation to known transcripts.

RNA turnover is an important way of controlling gene expression. While the characterization of the pathways and enzymes for RNA degradation are well-advanced in Escherichia coli and yeast, studies in Gram-positive bacteria have lagged behind. This tiling array study shows that two essential enzymes, the single-strand specific endonuclease RNase Y and the 5′–3′ exoribonuclease RNase J1, play central roles in the degradation of mRNAs in Bacillus subtilis. The double-strand specific enzyme RNase III plays a more minor role in RNA turnover, but has significant indirect effects on transcription. Depleting cells of these key enzymes led to the stabilization of many potentially regulatory RNAs, which were otherwise revealed only through testing a wide variety of experimental conditions. It is now possible to tell at a glance which of these three RNases is involved in the turnover of your favorite mRNA.

A hidden Markov support vector machine framework incorporating profile geometry learning for identifying microbial RNA in tiling array data

MOTIVATION

RNA expression signals detected by high-density genomic tiling microarrays contain comprehensive transcriptomic information of the target organism. Current methods for determining the RNA transcription units are still computation intense and lack the discriminative power. This article describes an efficient and accurate methodology to reveal complicated transcriptional architecture, including small regulatory RNAs, in microbial transcriptome profiles.

RESULTS

Normalized microarray data were first subject to support vector regression to estimate the profile tendency by reducing noise interruption. A hybrid supervised machine learning algorithm, hidden Markov support vector machines, was then used to classify the underlying state of each probe to 'expression' or 'silence' with the assumption that the consecutive state sequence was a heterogeneous Markov chain. For model construction, we introduced a profile geometry learning method to construct the feature vectors, which considered both intensity profiles and changes of intensities over the probe spacing. Also, a robust strategy was used to dynamically evaluate and select the training set based only on prior computer gene annotation. The algorithm performed better than other methods in accuracy on simulated data, especially for small expressed regions with lower (<1) SNR (signal-to-noise ratio), hence more sensitive for detecting small RNAs.

AVAILABILITY AND IMPLEMENTATION

Detail implementation steps of the algorithm and the complete result of the transcriptome analysis for a microbial genome Porphyromonas gingivalis W83 can be viewed at http://bioinformatics.forsyth.org/mtd.